The Nagssugtoqidian orogen in South-East Greenland: Evidence for Paleoproterozoic collision and plate assembly

Data Acquisition and Interpretation

Zircons from nine samples were dated by the SHRIMP U/Pb zircon method, using the SHRIMP 1 and SHRIMP RG instruments at the Research School of Earth Sciences, Australian National University and the SHRIMP 2 instrument at the Chinese Academy of Geological Sciences (table 1). ²³⁸U/²⁰⁶Pb in the unknowns was referenced to QGNG or FC1 zircons with ages of 1850 and 1099 Ma respectively, and U was calibrated against fragments of the 238 ppm U zircon SL13. Samples GGU–268867, GGU–232884, GGU–337136, GGU–337174 and EG86–66 were analyzed in the mid 1990s, prior to CL imaging being used routinely to document the interior structure of the zircons. Post-analysis CL imaging was undertaken on these in the 2000s, with additional CL-guided analyses undertaken in some samples. In addition, entirely CL-guided analyses were then undertaken on GGU–318487, GGU–337205 and GGU–337034. Analytical protocols, calibration of data and calculation of analytical errors are summarized by Stern (1998) and Williams (1998). Data assessment via Terra Wasserburg ²³⁸U/²⁰⁶Pb versus ²⁰⁷Pb/²⁰⁶Pb concordia plots, cumulative probability figures, weighted mean dates (reported in the text with 95% confidence limits) and multiple age components were calculated using the unmix routine of the Program ISOPLOT3.xla (Ludwig, 2003).

In situ trace element abundances in zircon (table 2) were obtained at RSES, ANU, using a Lambda Physik LPX 1201 UV ArF excimer laser instrument coupled to an Aligent 7500 inductively coupled plasma mass spectrometer (LA-ICP-MS). Sites for laser ablation analyses were chosen using the CL images as a guide and many were started over c. 1 μm deep SHRIMP U/Pb dating analytical pits. Data were assessed by plotting them as chondrite-normalized REE pattern diagrams. The zircons mounted in epoxy resin were placed in an Ar-He flushed sample cell (Eggins and others, 1998). The laser was operated at constant voltage, 21-23 kV at 5 Hz, with a spot diameter of 29 to 50 μm. Ablated material was carried by the He-Ar gas via a custom-made flow homogeniser to the ICP-MS. Data were acquired for c. 20 s with the laser beam turned off (for backgrounds) and c. 40 s with the laser on. This acquisition mode gives approximately 120 mass scans with a penetration depth of 20 μm. Each block of four unknowns was bracketed by analyses of glass reference material NIST610. Raw count rates were converted offline into concentrations using an Excel™ spreadsheet algorithm (written by S. Eggins and C. Allen, ANU). Corrections for mass bias drift in unknowns were made using the bracketing NIST610 glass analyses (Norman and others, 1996). Trace element abundances were normalized assuming Zr₂O = 64 wt% in the zircons (appropriate for a typical zircon with a few percent Hafnon solid solution).

Northern Archean Orthogneisses

Three samples of orthogneiss collected by the late David Bridgwater and others during the late 1970s helicopter reconnaissance mapping are representative of the variably retrogressed northern granulite facies gneisses (see fig. 1). The samples (GGU−268867a,b,c) are from a locality in the mountainous region Charcot Fjelde at approximately 66°24'N 36°59'W. Previous TIMS U-Pb zircon data (on different size and magnetic fractions from all three samples) have yielded an upper intercept age of 2835⁺⁶/_-8 Ma (Kalsbeek and others, 1993).

Zircons in these samples are mostly prismatic and up to 200 to 300 μm long, brown prisms dominated by oscillatory zoned domains, with variably developed mantles of homogeneous to sector zoned zircon. In some cases these form only narrow skins over the oscillatory-zoned zircon, whereas in other grains the mantles are broader and amenable to analysis using a ca. 30 μm SHRIMP spot. Twenty analyses were undertaken on zircons from GGU–268867a, five analyses on GGU–268867b zircons and thirteen analyses on GGU–268867c zircons (table 1 and fig. 4). In GGU–268867a multiple analyses of the outer rims of zircons visible by transmitted light microscopy were undertaken, in order to detect if any high-grade metamorphism affecting these rocks occurred in the Paleoproterozoic. However all such analyses yielded Neoarchean ages (table 1). Eight low- Th/U mantles in sample GGU–268867a yielded a spread in ages, but five gave a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2720 ± 6 Ma (MSWD=1.01), two gave older ages close to the age of the oscillatory zoned zircon and one gave an age of circa 2650 Ma. Interior oscillatory zoned zircon domains all yielded Neoarchean ages. In samples GGU–268867a oscillatory zircon is 2756 ± 6 Ma (n=11, MSWD=0.88, with one rejection) and in GGU–268867b 2734 ± 34 Ma (n=5; MSWD=3.3), which are interpreted as the age of the igneous protoliths of these gneisses. The oscillatory zoned zircon in sample GGU–268867c is some 100 Ma older. If the spread of measured ²⁰⁷Pb/²⁰⁷Pb in the zircons from this sample is attributed to some loss of radiogenic Pb during a circa 2720 Ma metamorphic event, then the presumed least disturbed analyses with the highest ²⁰⁷Pb/²⁰⁶Pb yield a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2863 ± 11 Ma (n=5; MSWD=0.89), which is interpreted as the minimum age of the igneous protolith of this gneiss (even the least disturbed zircons may have lost some radiogenic Pb). The results from these closely spaced samples support previous conclusions that (1) the Archean orthogneisses are polyphase in nature, with rocks of different ages occurring side by side, and (2) that at ca. 2720 Ma the granulite facies metamorphism in the northern fringe of the orogen is Neoarchean, rather than Paleoproterozoic (Kalsbeek and others, 1993).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Tera-Wasserburg diagrams of SHRIMP U/Pb zircon data from northern boundary zone orthogneisses GGU−268867a, b and c (A, B and C respectively). Analytical errors are depicted at 2 sigma. Inset on the right hand side of each frame shows the cumulate frequency distribution for ²⁰⁷Pb/²⁰⁶Pb (after correction for common Pb).

Orthogneisses Previously Termed ‘Blokken gneisses'

Sample GGU–232884a (65°53.4'N, 36°47'W) from the previously termed ‘Blokken gneisses' has a Nd T_DM value of 3010 Ma, suggesting predominance of Archean crustal components, and zircon dating of GGU–232884c and have yielded discordant ages with best-fit upper intersections at circa 2920 and 2960 Ma, respectively (Kalsbeek and others, 1993). Lower intercepts with Concordia were circa 1470 and 1900 Ma, both with a wide margin of uncertainty. Thus the previously termed ‘Blokken gneisses' are almost certainly Archean in age, whereas the bulk zircon lower intercepts with Concordia indicate new Paleoproterozoic zircon growth or Pb-loss from Archean zircon. The protolith age of one previously termed ‘Blokken gneiss' and its metamorphic history were examined by dating zircons from sample GGU–232884e by SHRIMP.

GGU–232884e yielded abundant 150 to 300 μm prismatic zircons dominated by weakly oscillatory domains, with the layering concordant to grain boundaries (fig. 5A). Irregular recrystallization domains overprint the oscillatory zoning. Most grains have CL-bright mantles (low U content), that generally are too narrow (<5 μm) for analysis by SHRIMP. All analyses yielded Archean ages, although all rims that were wide enough were analyzed. The oscillatory zoned zircon cores and whole grains yielded a spread of ²⁰⁷Pb/²⁰⁶Pb ages (fig. 6A), which is interpreted as a result of partial loss of radiogenic Pb in an ancient event. Accepting this model, the eight sites with the oldest ²⁰⁷Pb/²⁰⁶Pb ages and with an MSWD <1 yielded a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 3035 ± 14 Ma. This is interpreted as close to the age of the igneous protolith. Analyses of the low Th/U, low U rims that are CL-bright yielded Archean ages. Those with low common Pb content yielded a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 2723 ± 49 Ma (n=5; MSWD=0.07). This is interpreted as giving the time of zircon growth/recrystallization during a Neoarchean high-grade (granulite facies?) metamorphic event. Evidently, the Paleoproterozoic disturbance of the U/Pb isotope systematics of the zircons seen in the bulk zircon results (Kalsbeek and others, 1993) is restricted to some loss of radiogenic Pb, rather than new zircon growth.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Cathodoluminescence images of representative zircons from (A) previously-termed Blokken gneiss sample GGU−232884e, (B) Metatonalite GGU-337174, (C) Ammassalik Intrusive Complex diorite GGU−337205.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 6.

Tera-Wasserburg diagrams of SHRIMP U/Pb zircon data from (A) Blokken gneiss sample GGU−232884e. Multigrain IDTIMS data (Kalsbeek and others, 1993) is plotted with small circles. (B) Metatonalite 337174. (C) Ammassalik intrusive complex diorite GGU−337205. Analytical errors are depicted at 2 sigma. Inset on the right hand side of each frame shows the cumulate frequency distribution for ²⁰⁷Pb/²⁰⁶Pb (after correction for common Pb).

Paleoproterozoic Meta-tonalites

GGU–337174 is from an outcrop of garnetiferous tonalitic rocks west of Sermilik (66°06.8'N, 38°04'W). Bulk zircon fractions from samples 337169 and 337174 of these tonalites have yielded slightly discordant Paleoproterozoic ages, 1922 and 1850 Ma, and T_DM values of 2170 and 2190 Ma have been obtained from other samples of the same locality (Kalsbeek and others, 1993).

Zircons in GGU–337174 contain kernels of oscillatory-zoned zircon with often broad mantles that appear brighter and more homogeneous than the kernels in CL images (fig. 5B). The oscillatory-zoned kernels have high Th/U values and near concordant Paleoproterozoic ages (table 1, fig. 6B). Assuming the spread in ²⁰⁷Pb/²⁰⁶Pb beyond analytical error of these data to be the product of some ancient loss of radiogenic Pb, a group with the highest indistinguishable ²⁰⁷Pb/²⁰⁶Pb yielded a weighted mean age of 1901 ± 9 Ma (n=19; MSWD=0.64, one reject). This is interpreted as the formation age of the tonalite protolith to this gneiss. The mantles and replacement domains that appear bright and homogeneous in CL images are lower in U with somewhat lower Th/U (table 1). Four out of five analyses of such sites yielded a weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 1840 ± 56 Ma (MSWD=0.36). This indicates that zircons in this rock were recrystallized late in the magmatic event or during a high-grade metamorphism shortly afterwards.

LA-ICP-MS trace element analyses were undertaken on the zircons in order to gain further information on zircon recrystallization events (table 2). All analyses display negative Eu anomalies (fig. 7A) indicating equilibration with plagioclase in both igneous zircon growths and metamorphic recrystallization. Both oscillatory-zoned kernels and the more homogeneous grain margins show strong enrichment of the HREE over the LREE. The degree of HREE enrichment in some recrystallization domains is as high as that in some igneous domains, despite the tendency for trace elements to be expelled when zircons recrystallize. This process did not involve a change in Gd/Lu_(N) of the zircons (fig. 7A), showing no change in the role of garnet as a coexisting phase during the crystallization history of the zircons. The trace element data demonstrate that the sporadic occurrence of garnet in this suite did not influence the REE spectra in the zircons.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 7.

Chondrite normalized plots of LA-ICP-MS REE analyses. (A) zircons from Paleoproterozoic tonalite 337174. (B) metamorphic zircons from eclogite facies domain in Paleoproterozoic metadiabase dike 318487.

Ammassalik Intrusive Complex Diorite

GGU–337205 is a diorite from the Ammassalik Intrusive Complex. It was taken from a unit of homogeneous undeformed orthopyroxene-bearing diorite from the quarry in Ammassalik town (65°36.5'N, 37°38'W). Bulk zircon analyses of this sample yielded a close to concordant Paleoproterozoic age, with a Concordia intercept age of 1886 ± 2 Ma (Hansen and Kalsbeek, 1989). The zircons from GGU–337205 are large (200 –500 μm) prisms and prism fragments, with widely developed oscillatory-zoning. However, this zoning is disrupted by homogeneous to blotchy-textured domains that appear bright in CL images (fig. 5C). All analyses yielded close to concordant Paleoproterozoic dates (table 1, fig. 6C). All oscillatory-zoned zircon analyses yielded a weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 1881 ± 10 Ma (n=17; MSWD=0.92) that is interpreted as the age of intrusion. With one analysis rejected, those of the bright in CL domains have lower U contents and yielded a weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 1905 ± 49 Ma (n=7; MSWD=0.79), which within its large uncertainty is indistinguishable from the age of igneous crystallization.

Provenance and Metamorphic History of Sillimanite and Orthopyroxene Bearing Metasedimentary Rocks Associated with the Ammassalik Intrusive Complex

Garnet + sillimanite + orthopyroxene bearing metasedimentary rocks along the southern edge of the Ammassalik Intrusive Complex (fig. 1) contain extensive garnet-bearing granitic melt domains that migrate and mingle with the diorites of the Complex. Thus from field relationships it appears that melting in the metasediments was triggered by emplacement of the Ammassalik Intrusive Complex. IDTIMS analyses of bulk zircon fractions from garnet-granite gneiss GGU–337041 yielded an upper intersection with Concordia at 1900⁺²⁵_-35 Ma and a lower intersection at ∼zero, with ²⁰⁷Pb/²⁰⁶Pb ages between 1913 and 1903 Ma (Kalsbeek and others, 1993). This age agrees with those obtained from the Ammassalik Intrusive Complex and supports the field interpretation that melting of metasedimentary gneisses was triggered by intrusion of the complex.

Provenance of the metasedimentary rocks and their metamorphic history was examined further by zircon dating of diatectic garnetiferous metasediment GGU–337034. In CL images the zircons often display complex structure, with more than one structural domain (fig. 8A). Rounded and truncated oscillatory-zoned cores are obvious in some grains, whereas in others a kernel of oscillatory zoned zircon with zoning more or less parallel to grain boundaries is mantled by domains of more homogeneous zircon. There are also some equant, structureless grains. Fifty analyses were undertaken of all types of zircon domain, with thirty-nine grains being studied (table 1, fig. 9A). Only one Archean core (grain 24) was detected, with all other zircon yielding Paleoproterozoic ages, mostly of 1950 to 1850 Ma, with only a few giving slightly older ages (for example, grain 6 with an age of circa 2050 Ma). For zircon of undoubted detrital origin that forms rounded cores in grains with truncated oscillatory zoning (for example, grain 6, fig. 8A) youngest ages were circa 1910 Ma (for example, analyses 35.1 and 36.1). Outermost mantles on grains are generally least structured in CL images (for example, the protuberance on grain 22, fig. 8A). There are also a few equant grains that are structureless and CL-dull (grain 13, fig. 8A). These domains generally have moderate to low Th/U. They yield a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 1874 ± 8 Ma (95% confidence, MSWD=0.41). This is taken to mark zircon growth and recrystallization during a high grade metamorphic event. It should be noted that this age population does not contain any oscillatory-zoned zircon. There are also other abundant zircon domains, which structurally appear older than the 1874 ± Ma metamorphic domains, yet do not form obvious rounded >1910 Ma cores. This zircon shows a range of morphology and Th/U. Some domains (for example the main part of grain 22, fig. 8A) are oscillatory zoned, with zoning parallel to grain margins. Other domains are homogeneous generally CL-bright (low U), and can form rims or recrystallization domains (for example on grain 23, fig. 8A). These domains are intermediate in age between the 1874 ± 9 Ma rims and equant grains, and cores definitely of detrital origin. We interpret this complex suite of zircon domains to have grown during the diatexis event that affected this rock. We contend that the unrounded oscillatory-zoned zircons grew in neosome melt domains, and the structureless rims and equant grains grew sub-solidus in paleosome. This age is indistinguishable from that of the Ammassalik Intrusive Complex. Thus it appears that the sedimentary protolith of this rock carried mostly Paleoproterozoic detrital zircons, with only rare Archean grains. It was affected by diatexis at the time the Ammassalik Intrusive Complex was intruded, and then by further metamorphism, approximately 25 million years later.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 8.

Cathodoluminescence images of representative zircons from (A) diatectic metasedimentary gneiss GGU−337034 associated with the Ammassalik intrusive complex and (B) kyanite-bearing metasediment GGU−337136.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 9.

Tera-Wasserburg diagram of SHRIMP U/Pb zircon data from kyanite-bearing metasediment GGU−337136. Analytical errors are depicted at 2 sigma. Inset on the right hand side of the frame shows the cumulate frequency distribution for ²⁰⁷Pb/²⁰⁶Pb (after correction for common Pb).

Provenance and Metamorphic History of Kyanite-Bearing Metasedimentary Rocks Intercalated with Archean Gneisses

GGU–337136 is a garnet- and kyanite-bearing paragneiss from a circa 200 m wide layer of metasedimentary rocks embedded in orthogneisses of probable Archean age at Augpalugtooq (66°06.5'N, 37°57'W; fig. 1). The relationship with the surrounding orthogneisses is ambiguous due to high strain. Sm-Nd data on two paragneisses yielded T_DM values of 2.88 and 2.75 Ga indicating that they consist mainly of Archean detritus. However, Rb-Sr data on 14 samples yielded a well constrained errorchron with a slope corresponding to an age of 1870 ± 50 Ma, and the relatively low initial ⁸⁷Sr/⁸⁶Sr ratio of 0.714 strongly suggests that the present high Rb/Sr ratios of the samples (mean Rb/Sr = 0.94 –much higher than that of the surrounding gneisses) are probably related to sedimentary processes, and suggest Paleoproterozoic deposition (Kalsbeek and others, 1993).

Dating of zircons from GGU–337136 was undertaken in 1994 without assistance of CL images. However, images were later obtained on some of the previously analyzed grains to aid data interpretation. GGU–337136 yielded abundant small zircons, which reveal complex structure in CL images (fig. 8B). Centers of grains contain oscillatory-zoned to homogeneous zircon. These are partly replaced and/or overgrown by domains that in CL imaging reveal sector zoning or appear homogeneous. These new domains can themselves show complex internal structure (fig. 8B), and might reflect more than one episode of zircon recrystallization and growth. The older oscillatory-zoned zircon can show truncated zonation at grain boundaries, and is interpreted to be detrital in origin. Eight analyses of detrital zircon yielded discordant Paleoproterozoic ages (table 1, fig. 9B), and are interpreted as Archean grains that lost some radiogenic Pb in the Proterozoic. Analyses of twenty-two other detrital zircons yielded Archean ages between 2800 and 2600 Ma. Thus the maximum depositional age of the sediment is around 2600 Ma. The recrystallization and overgrowth domains generally have low Th/U (table 1) and have close to concordant Paleoproterozoic ages. These domains show a dispersion of ²⁰⁷Pb/²⁰⁶Pb ages beyond that expected from a single population, in keeping with their complex structure in CL images. Using the unmix routine of Isoplot, three age components of 1740 ± 40 (9%), 1835 ± 19 (59%) and 1874 ± 23 Ma (32%) were extracted. Although the age of any of these components might be open to debate, clearly the zircons suffered complex recrystallization and regrowth in the Paleoproterozoic. One age component (1740 ± 40 Ma) is similar to a Pb/Pb whole rock isochron age (1773 ± 22 Ma) obtained from marbles from the northern part of the mobile belt by Taylor and Kalsbeek (1990). The zircon geochronology places the deposition of the sediment between c. 2600 and c. 1900 Ma. Therefore it is unlikely to be an integral component of the Neoarchean gneiss complex, but instead belongs to an early Paleoproterozoic cover sequence.

Age and Metamorphic History of Paleoproterozoic Metadiabase Dikes

Sample EG86–66 is a fractionated felsic domain in a metadiabase dike from the northern coastal part of the orogen, and was collected by the late David Bridgwater (fig. 1). Regrettably, with David's death, more detailed field information on this sample is not available. It gave a small yield of prismatic to jagged anhedral zircons (fig. 10A). This habit is typical of zircons in gabbros (Poldervaart, 1956), and probably indicates growth in late interstitial melt domains of local zircon saturation. In CL images the zircons contain interior domains that appear dull, but locally display weak oscillatory zoning. These are partially replaced and overgrown at their margins by domains that are bright and homogeneous or patchy in CL images (fig. 10A). The interior domains are interpreted to be of igneous origin, whilst the bright exterior domains are interpreted to have developed by recrystallization and perhaps regrowth during metamorphism. Analyses of CL-dull domains have moderate to high U content (1100 –300 ppm) and high Th/U (0.65 –0.4), with a spread of ²⁰⁷Pb/²⁰⁶Pb ages between circa 2020 Ma and 1800 Ma (table 1, fig. 11A). Multiple analyses on some of the grains suggest this spread is due to variable partial ancient loss of radiogenic Pb from one generation of zircon. Three analyses with the oldest ²⁰⁷Pb/²⁰⁶Pb yield a weighted mean date of 2015 ± 15 Ma (MSWD=0.69). This is interpreted to approach the time of intrusion of the metadiabase dike. All the CL-bright domains have lower U concentrations and yield a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 1836 ± 58 Ma (n=5; MSWD=0.62). This is interpreted as the time of zircon recrystallization during metamorphism.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 10.

Cathodoluminescence images of zircons from metadiabase dike samples (A) EG86−66 and (B) GGU−318487.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 11.

Tera-Wasserburg diagrams of SHRIMP zircon U/Pb data from metadiabase dike samples (A) EG86−66 and (B) GGU−318487. Analytical errors are depicted at 2 sigma. Inset on the right hand side of each frame shows the cumulate frequency distribution for ²⁰⁷Pb/²⁰⁶Pb (after correction for common Pb).

Sample GGU–318487 of an eclogite facies domain in a dike north of the Ammassalik Intrusive Complex (fig. 1) yielded small equant zircons that appear homogeneous to sector-zoned in CL images (fig. 10B). The exterior of some grains have thin domains that appear darker in CL images. These were too narrow to analyze with a typical >20 μm SHRIMP spot. A few grains contain small, corroded domains that appear dull in CL images. One SHRIMP analysis attempted on such a domain had the highest U content (1181 ppm) and highest Th/U (0.27) with a ²⁰⁷Pb/²⁰⁶Pb date of circa 1900 Ma (grain 9; table 1, fig. 11B). This is interpreted as a small (disturbed) relict of igneous zircon, showing the dike is at least 1900 Ma old. Analyses of the homogeneous to sector zoned zircon (table 1; fig. 11B) have rather low U concentrations (mostly <70 ppm) and low Th/U (mostly <0.03) and with rejection of one analysis with an anomalously young age yielded a weighted mean ²⁰⁷Pb/²⁰⁶Pb date of 1867 ± 28 Ma (n=13; MSWD=0.60). The dominant homogeneous to sector-zoned zircons contain rare micro-inclusions of quartz, clinopyroxene and Ca-garnet, with no plagioclase inclusions found. LA-ICP-MS trace element analyses of these zircons reveal depressed HREE abundances with no significant negative Eu anomalies (table 2, fig. 7B). This indicates zircon growth in a garnet-rich but plagioclase-absent environment (Rubatto, 2002). Thus 1867 ± 28 Ma is interpreted as the time of eclogite facies metamorphism (plagioclase absent), rather than that of growth during subsequent isothermal recrystallization at lower pressures (plagioclase present).

Kalsbeek and others (1993) reported a Sm-Nd garnet-clinopyroxene-whole rock isochron of 1817 ± 22 Ma for an eclogite remnant in a dike, marginally younger than the 1867 ± 28 Ma date obtained for the eclogite facies metamorphism here. Petrography of the eclogites indicates high temperature decompression reactions, such as exsolution of albite from originally omphacitic clinopyroxene and the development of orthopyroxene + plagioclase on the edges of garnets (Nutman and Friend, 1989). Thus the 1817 ± 22 Ma Sm-Nd mineral isochron is considered to reflect closure of minerals to Sm and Nd diffusion along a high temperature decompression path, rather than to date maximum pressure of metamorphism.

Reassessment of the Nagssugtoqidian Orogenic Belt in East Greenland

The results presented here provide new insights into the history of the Nagssugtoqidian orogen in the Ammassalik region (fig. 12), and support some previous interpretations, but necessitate revision of others. The zircon geochronological results combined with the broader isotopic studies of Kalsbeek and others (1993) confirm that Archean and Paleoproterozoic rocks with contrasting origins and metamorphic histories are present. The new results, particularly the focus on dating of metamorphic rims and recrystallization domains in zircons provides more concrete information on a long and diverse metamorphic history. Thus for Archean rocks in the northern part of the region, circa 2720 Ma is shown to be the time of high-grade (granulite facies) metamorphism. This study reinforces previous work (Kalsbeek and others, 1993) that Archean gneisses in the north of the region (affected by 2720 Ma metamorphism) show different protolith ages (3035−2750 Ma). The Ammassalik Intrusive Complex, which is dominated by juvenile additions to the crust, experienced 1905 ± 49 Ma LP-HT metamorphism up to granulite facies, probably coeval with its igneous emplacement. On the other hand, structurally underlying Archean gneiss complexes have Paleoproterozoic metadiabase dike remnants that display eclogite facies metamorphism at 1867 ± 28 Ma (U/Pb zircon) with closure to Sm-Nd diffusion along a high temperature decompression path at 1817 ± 22 Ma (Sm-Nd mineral isochron).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 12.

Paleoproterozoic P-T-t (pressure-temperature-time) evolution in the Ammassalik region, based on results presented here and by Kalsbeek and others (1993), combined with reassessment of data in Nutman and Friend (1989).

Previously, Friend and Nutman (1989) presented a P-T-t synthesis where the emplacement of the Ammassalik Intrusive Complex and its low-pressure metamorphism were younger than the eclogite facies metamorphism. This followed on from the interpretation of Bridgwater and Myers (1979) that the Ammassalik Intrusive Complex is a late-kinematic intrusion. The accumulated zircon dating now suggest that such a scenario is unlikely, because the Ammassalik Intrusive Complex (1886 ± 2 Ma) appears to be older than the eclogite event (1867 ± 28 Ma) –although at the two sigma level the error intervals overlap, the probability of a real age difference is> 90 percent. Also, by comparison with similar rocks in the Nagssugtoqidian belt of West Greenland for which an arc setting is indicated by geochemical data, we suggest that the Ammassalik Intrusive Complex is a magmatic arc complex dominated by juvenile Paleoproterozoic material (Kalsbeek and others, 1993). This is in accord with data presented here that shows the importance of Paleoproterozoic (2100−1900 Ma) rather than Archaean zircons. It is proposed here that the HT-LP metamorphism in the Ammassalik Intrusive Complex is unique to those rocks, and that it along with associated supracrustal rocks were tectonically emplaced over the edge of a southern terrane of Archean continental crust and cover sequences, which then experienced transient high pressure metamorphism at 1867 ± 28 Ma (fig. 12). This is consistent with southerly−directed tectonic transport deduced in earlier studies (Chadwick and others, 1989). This tectonic emplacement/collision led to the cessation of arc magmatism. Subsequently the Paleoproterozoic juvenile arc rocks, cover sequence and southern Archean basement experienced a common history of folding and amphibolite facies metamorphism (fig. 12). Such a model infers that subduction was northward, meaning that the Ammassalik Intrusive Complex would be a component of an overlying arc. This would place a suture at the structural base of the Ammassalik Intrusive Complex and its carapace of metasedimentary rocks with HT-LP (orthopyroxene grade) metamorphism. This hypothetical suture would not have a simple geometry, because it has been folded at least twice and has entirely or partly been excised by younger shear zones, such as the steeply dipping one at the northern edge of the Complex (fig. 1). This reappraisal supports previous suggestions (Kalsbeek and Taylor, 1989) that the Nagssugtoqidian orogen in East Greenland should contain a (cryptic) suture. Moreover, the occurrence of almost coeval HP and LP metamorphic complexes is a characteristic of Phanerozoic convergent plate boundary processes (Miyashiro, 1973), supporting the Paleoproterozoic plate tectonic scenario explored here for the Ammassalik region.

Comparison of Events of the Nagssugtoqidian in East and West Greenland

In the 1990s, following on from the geochronological and geochemical study of Kalsbeek and others (1987), a major research initiative in the Nagssugtoqidian of West Greenland found sheared sutures (or one suture repeated by thrusting) between unrelated blocks of Archean basement crust and juvenile Paleoproterozoic arc assemblages (Kalsbeek and Nutman, 1996; van Gool and others, 2002). Where the suture had not been excised by later shear zones, they are commonly marked by metasedimentary rich units rich in marbles (for example, Kalsbeek and Nutman, 1996). These would have made excellent decollements during thrust transport, which has been estimated to be more than 50 km (Kalsbeek and Nutman, 1996; Manatschal and others, 1998; van Gool and others, 2002). The fundamental nature of these tectonic contacts is illustrated by the common occurrence of peridotite pods of likely mantle origin along them (Kalsbeek and Manatschal, 1999). Detailed zircon geochronology showed that arc magmatism lasted from 1920 to 1870 Ma, that the youngest detrital zircons in immature quartzo-feldspathic sediments in the arc assemblages are circa 1900 Ma old, and that regional metamorphism of all units peaked at 1850 to 1840 Ma (Taylor and Kalsbeek, 1990; Kalsbeek and Nutman, 1996; Whitehouse and others, 1998; Nutman and others, 1999; Connelly and others, 2000). Ar-Ar studies show that this was followed by a long period of cooling (Willigers and others, 2001, 2002). Hence there are similarities between the chronology of arc magmatism and metamorphism in the eastern and western segments of the Nagssugtoqidian (fig. 13). This is compatible with geophysical evidence that they are the same orogenic belt (Verhoef and others, 1996).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 13.

Synthesis of Paleoproterozoic crustal evolution of Greenland. (A) Delineation of northern, central and southern juvenile arcs within Paleoproterozoic orogenic belts. (B) Ages for arc magmatism, youngest detrital zircons in arc-related rocks, metamorphism, post collisional granites, age of mafic dikes, and extensional basins. See text for explanation and key references.

Assembly of Greenland from Paleoproterozoic Arcs and Archean ‘Microcontinents'

Kalsbeek (1981) proposed on the basis of Rb-Sr whole rock data that Archean basement was widespread north of the Paleoproterozoic Nagssugtoqidian front along the northern edge of the North Atlantic Craton (fig. 13). The wide scatter about individual Rb-Sr isochrons indicated that in many areas Archean basement gneisses had been affected by superimposed Paleoproterozoic (‘Hudsonian') metamorphism. The most obvious exception was the Ketilidian orogen of South Greenland, which by then was already recognized to contain Paleoproterozoic plutonic rocks (for example, van Breemen and others, 1974).

In the 1980s there was broader application of zircon and Sm-Nd whole rock geochronology to basement complexes. In the Nagssugtoqidian orogenic belt in West Greenland, these methods identified tracts of Paleoproterozoic meta-granitoids and gneisses with radiogenic isotopic signatures showing that they are juvenile additions to the crust, and whose calc-alkaline geochemistry resembles that found in Phanerozoic arcs at convergent plate boundaries (Kalsbeek and others, 1987). These are embedded in ‘reworked' Archean gneisses of the Nagssugtoqidian orogenic belt. From this data, Kalsbeek and others deduced that north of the Ketilidian orogen, the gneissose basement of Greenland (starting with the Nagssugtoqidian mobile belt), contains cryptic Paleoproterozoic sutures marked by rocks formed in Paleoproterozoic arcs.

By the 1990s with copious zircon U/Pb dating (particularly by SHRIMP), Nd isotopic analysis and further methodical mapping, a major tract of juvenile Paleoproterozoic crust was found to form much of the basement in the Caledonian orogen of North-East Greenland (Kalsbeek and others, 1993) and within the Inglefield mobile belt of Northern West Greenland (Dawes and others, 1987; Nutman and others, 2008). Correlation between these Paleoproterozoic domains in northern Greenland is uncertain, and it is possible that the two belts are entirely unrelated.

With the completion of a geochronological program in the Inglefield belt of North-West Greenland (Nutman and others, 2008) together with the results from the eastern sector of the Nagssugtoqidian belt presented in this paper, it is possible to make the most complete synthesis yet of the evolution of the Greenland basement (fig. 13). This synthesis is an amalgamation of numerous geological and isotopic studies, for which only key references can be given here. This synthesis extends, and largely agrees with, the recent one by St-Onge and others (2006). Three tracts of Paleoproterozoic juvenile arc rocks have been recognized in Greenland, particularly on the basis of zircon dating and Nd isotopic studies (for example, Kalsbeek, 1986; Kalsbeek and others, 1993a, 1993b; Whitehouse and others, 1998; Garde and others, 2002 and references therein).

The northern arc tract (fig. 13) dominates the Inglefield mobile belt of North-West Greenland. The extension of this arc underneath the Inland Ice to rocks of similar age and origin on the east coast is unconfirmed. The northern boundary of this arc is largely obscured by the Inland Ice and by Mesoproterozoic-Paleozoic platform sediments. However a window to the basement in Victoria Fjord (VF on fig. 13) contains Archean basement gneisses (Hansen and others, 1987; Nutman and others, 2008) and Sm-Nd data on erratic granitic boulders derived from crystalline basement underneath the Inland Ice further east yield Archean T_DM values (Kalsbeek and Frei, 2006). These occurrences indicate the presence of Archean basement in northernmost Greenland. The northern arc tract in North−West Greenland is bounded on its southern side by intercalations of quartzites, pelites and marbles with Archean gneisses that were metamorphosed and deformed in the Paleoproterozoic (Dawes and others, 1987; Nutman and others, 2008). Metamorphosed diabase dikes in the northernmost Archean basement contain garnet (largely decompressed to plagioclase and amphibole; Nutman, 1984) and in Archean metasediments, (Archean?) orthopyroxene is mantled by garnet corona structures (Garde and others, 1984). These might point to transient early high pressures during Paleoproterozoic metamorphism.

In West Greenland between the northern and central arcs lies the Rinkian mobile belt (Henderson and Pulvertaft, 1987), in which Archean basement gneisses and overlying Paleoproterozoic metasediments (Kalsbeek and others, 1998 and references therein) were intensely folded into large-scale recumbent structures. Early high-grade metamorphism has been dated at 1881 ± 20 Ma (Taylor and Kalsbeek, 1990). Basement gneisses and their Paleoproterozoic cover were intruded by the 1869 ± 9 Ma Prøven Charnockite (fig. 13; Thrane and others, 2005, and references therein). The Prøven Charnockite is a large, orthopyroxene-bearing granite with geochemical affinity with A-type granites, rather than calc-alkaline juvenile arc suites (Thrane and others, 2005). Thrane and others interpreted the Prøven Charnockite to have formed in the passive margin side of a colliding pair of Archean cratons. It might form the edge of the massive Cumberland batholith of the same age to the west in Arctic Canada (fig. 13; Thrane and others, 2005).

The central arc tract lies within the Nagssugtoqidian orogen (fig. 13). In both East and West Greenland, the Nagssugtoqidian mobile belt is bounded on its north side by domains of Archean crust that are devoid of penetrative Palaeoproterozoic deformation and high grade metamorphism (for example, this paper). The belt contains a complex collage of (a) Paleoproterozoic magmatic rocks with arc-like geochemical signatures intruded into immature quartzo-feldspathic sediments derived mostly from Palaeoproterozoic detritus (Kalsbeek and others, 1987; Nutman and others, 1999; Kalsbeek, 2001), (b) earliest Paleoproterozoic quartzites and marbles with abundant Archean detrital zircons and evidence of early high pressure metamorphism via kyanite and (c) Archean gneisses cut by variably, but generally strongly deformed Paleoproterozoic metadiabase dikes with vestiges of high pressure metamorphism (this paper and Nutman and others, 1999). The latter occur deep within the orogen, and in the northernmost autochthonous parts of the North Atlantic Craton (Nutman and others, 1999). Within the central part of the orogen, detailed mapping combined with zircon dating shows that a juvenile Palaeoproterozoic arc assemblage was thrust at least 50 km over Archean gneisses capped by marble-rich sediments that acted as the decollement (Kalsbeek and Nutman, 1996). A recent synthesis of the western segment of the orogen proposes two northerly-directed subduction zones, the southern of which gave rise to the Sisimiut charnockite complex where Nd isotopic data shows some contribution from Archean crust, and the northern one represented by the Arfersiorfik complex, with lesser contributions from Archean crust (St-Onge and others, 2006).

The southern arc tract lies within the Palaeoproterozoic Ketilidian orogen of southernmost Greenland. Detailed work in this orogen summarized by Garde and others (2002) demonstrates that it contains an Andean-style arc named the Julianehåb batholith, developed over a northerly-dipping oblique subduction zone at the southern edge of the North Atlantic Craton. A highly metamorphosed forearc sedimentary prism related to the Julianehåb batholith forms southernmost Greenland.

The Paleoproterozoic arc-related rocks, the youngest ages of detrital zircons in associated immature volcanosedimentary sediments, regional metamorphism, dating of high pressure metamorphism, younger syncollisional granites and post-orogenic granite intrusions are assembled in figure 13. These reveal that throughout much of Greenland there is a southerly progression in the ages of Palaeoproterozoic crustal accretion and metamorphism. The Palaeoproterozoic arc rocks in the Caledonian basement of North-East Greenland are presently left out of this appraisal, because it is uncertain how they link with rocks of similar character to the west on the other side of the Inland Ice. The northern arc tract contains 1950 to 1920 Ma juvenile crustal components, the youngest detrital zircons are circa 1980 Ma, and regional metamorphism occurred at circa 1920 Ma. Metamorphism in the arc assemblage was HT-LP (orthopyroxene grade). The central arc tract (in the Nagssugtoqidian), contains 1920 to 1870 Ma arc rocks. The detrital zircons in immature metasediments associated with the arc rocks are overwhelmingly Palaeoproterozoic, with the youngest being circa 1900 Ma old (Nutman and others, 1999). In the East Greenland segment of this orogen, high pressure metamorphism up to eclogite facies in underlying basement rocks has been documented at circa 1870 Ma (this paper) coeval with the youngest-dated intrusions (Connelly and others, 2000) from the arc assemblages in West Greenland. This probably marks the end of arc magmatism with the collision of northern and southern continental masses, crustal thickening and transient high pressure metamorphism at depth. All rocks of the central orogen show high-grade metamorphism at 1850 to 1840 Ma (for example, Taylor and Kalsbeek, 1990; Nutman and others, 1999; Connelly and others, 2000; this paper), coeval with folding and local anatexis during crustal equilibration. In the southern arc (Julianehåb batholith in the Ketilidian orogen) the onset of arc magmatism at the southern margin of the North Atlantic Craton was at circa 1840 Ma and lasted until circa 1790 Ma (Garde and others, 2002, and references therein). The youngest detrital zircons associated with the Julianehåb batholith are circa 1790 Ma. This was succeeded between 1790 to 1740 Ma by metamorphism and the intrusion of S-type granites formed by melting of the forearc sedimentary prism and by rapakivi granites in the southernmost part of the orogen.

Thus during the Palaeoproterozoic (2000−1790 Ma), Greenland basement was created by the growth of arc complexes and their collision and amalgamation with 2000 to 1000 km wide Archean basement gneiss massifs (‘microcontinents') fringed by Palaeoproterozoic passive margin sequences (clean quartzites, carbonates and pelites). For the southern arc it is well established that subduction was northwards towards the North Atlantic Craton, and that it entailed a degree of oblique movement. We suggest in this paper that subduction related to the Ammassalik region eastern section of the central arc tract was also northwards, such that upon collision, the arc assemblage overrode the edge of the North Atlantic Craton (in which transient high pressure metamorphism took place; fig. 13). This is in accord with suggested polarity for probably 2 subduction zones within the western section of the Nagssugtoqidian (for example, St-Onge and others, 2006). The presence of local Palaeoproterozoic high pressure granulite assemblages at the northern edge of the North Atlantic Craton would fit with northerly−directed subduction. For the northern arc, northwards subduction is inferred by Nutman and others (2008).

Across Greenland, latest Palaeoproterozoic events between 1775 and 1680 Ma were intrusion of A-type granites including rapikivi granites in the Ketilidian orogenic belt and transcurrent shearing. Although much of this activity is found within the Palaeoproterozoic mobile belts, Rb-Sr mineral ages demonstrate general heating and locally intrusion of granite dikes in the Archean domains (Kalsbeek and others, 1980; Baadsgaard and others, 1986). Unlike the timing of Palaeoproterozic arc magmatism and regional metamorphism, there are no changes in ages across Greenland for this activity (fig. 13). These pan-Greenland thermal event(s) might indicate the start of extension and high heat flow leading to continental break-up in the Mesoproterozoic.